Design and implementation of the forecasting system for wind farm power

With the large-scale wind farm in the Grid, the stability of power system is becoming the hot issue. It makes higher requirements to the accuracy of power put forward in the forecasting system. This paper is raised the forecasting model based on multi-Agent technology, given prediction algorithm module flow and used BP neural network to predict. According to the characteristics of self-learning Agent, the paper makes use of all interactivities of Agent and modifies the model of BP neural network constantly to make predictions more accurately. The result of actual system application show that the forecasting model based on multi-Agent technology is feasibility and effectiveness.